gitlab-links's Introduction

GitLab Link Styler

Styles closed issues and merge requests on GitLab. Reads links to issues, and looks each one up in GitLab via the GitLab API. Closed issues are crossed out. Open issues with checklists in their description are underlined with a progress bar.

In this example, the issues 21528 and 21613 were closed (at the time of the screenshot), while 21706 was open, and had a checklist in it's description, so a progress bar was added as an underline.

Installation

With TamperMonkey installed in your browser, visit the Raw Script, which TamperMonkey will prompt you to install.

The script is set to automatically update itself, but the TamperMonkey extension has a very slow update interval by default. I recommend adjusting the update frequency in the TamperMonkey settings.

Settings

Visit http://gitlab-links.codelenny.com/ to configure the plugin.

Change Log

v0.3 (8/11/2016)

Added assets forgotten in v0.2

v0.2 (8/11/2016)

Added _version flag to the stored data, to ensure bugfixes to caching are applied.
Fixed timestamp caching issue
Tweaked rounding:
- if the percent of completed tasks would round to 0%, but one or more tasks are completed, round up to 10% so the issue doesn't appear unstarted
- if the percent of completed tasks would round to 100%, but one or more tasks are not yet completed, round down to 90% so the issue doesn't appear fully resolved

gitlab-links's People

Contributors

Stargazers

Watchers

gitlab-links's Issues

Style Merge Request References

Extract data from issue page

Directly update the cache if the user changes an issue by parsing the page

Bulk fetch issues

Try to reduce load by not loading issues/merge requests individually. Not sure which will be best for GitLab.

Possible methods:

Individual requests (current)
Only pagination and caching: fetch the closest 20 (a single page) to an issue, in the hopes that one will be useful either immediately or in the next 20 minutes (before the cache expires)
Flexible pagination: if only a single issue/MR is requested from a project, use an individual request. Otherwise, figure out a balanced number of issues to fetch (by playing with the number of issues per page) to get 2-3 issues at once, and cache the issues in between.

Recommend Projects